This competing renewal will provide continuing support for the Syntheses of HIV/AIDS Research Project (SHARP). The studies will center on issues central to factors that underlie the efficacy of HIV risk-reduction interventions, with a series of studies conducted for each of three interrelated specific aims related to intervention content, maintenance of risk behavior change, and structural interventions for HIV prevention. First, intensive, psychosocial behavioral interventions will be content-analyzed across a broad set of HIV prevention studies and these dimensions will be related to the trials' efficacy. Structural aspects integrated into individual trials will also be evaluated. These results will inform not only key content dimensions but also highlight improvements to reporting requirements of future intervention trials. Second, theory-driven meta-analyses will examine how sexual risk-behavior change can best be maintained and with what success sexual infections are averted, focusing in turn on adolescents, heterosexual women, men who have sex with men, and African-Americans; each of these will consider the role of supportive social milieus in facilitating or hindering efficacy. Moreover, some will examine prevention across populations and the roles of HIV knowledge and levels of protected sexual interactions at the outset of trials. Third, meta-analyses will be performed of the literatures relating to prevention of HIV through changes in social and policy structure that affect the availability of preventative actions, including condom availability and condom-use policies needle exchange programs, and increases in services for sexually transmitted infections. By carefully summing up the evidence to date on these important and quickly growing literatures, these meta-analyses will inform public health officials, community-based interventionists, and scientists about which interventions work best, which components of interventions have the largest impact, and the circumstances under which HIV risk-reduction interventions are most likely to be produce lasting change.